Cox proportional hazards (pH) regression analyses were conducted to predict initiation of cannabis use and transition from cannabis initiation to onset of CUD symptoms. Age at first cannabis use was the entry point in the CUD symptom onset analyses. Cox pH regression analyses account for the possibility that participants who have not yet experienced the event of interest (e.g. initiation of cannabis use) may do so in the future. Under this approach, data up until the time of censoring (most recent interview) is used in the calculation of hazard ratios. The pH assumption that risk remains constant over time was tested with the Grambsch and Therneau test of the Schoenfeld residuals (Grambsch and Therneau, 1994), violations of the assumption were corrected with age specific interactions, which can reveal the extent to which risk associated with trauma varies across the period of risk (e.g., whether trauma exposure is a predictor at certain ages). Trauma exposures, PTSD, MDD, tobacco and alcohol involvement were modeled as time-varying covariates coded as negative in each year prior to the age at first occurrence and positive
